AUTHOR
Hostettler, Roland

DEPARTMENT
Computer Science and Electrical Engineering / Systems and Interaction

SUMMARY
In nowadays road traffic, road user safety is a main concern for vehicle manufacturers and road traffic administrations. New technologies enable complex and save vehicles and modern building techniques are the key to high quality roads. Intelligent road infrastructure systems are available, e.g. for adaptively routing traffic or surveillance of crucial traffic junctions. However, such systems are relatively sluggish in providing assistance to drivers of individual vehicles anywhere along the road.

Within the framework of the iRoad project, this thesis focuses on the exploitation of road surface vibrations for vehicle parameter estimation. Measured road vibrations are analyzed and phenomena are identified. Algorithms for vehicle detection and axle estimation based on the findings are derived and their performance is evaluated and discussed.

It is found that a vehicle detection of over 99% can be achieved by using the proposed algorithm, independent of the vehicle speed. Axle parameter estimation yields reasonable results for the number of axles and their relative positions as well as transferred energy. In combination with another method to determine the vehicle speed, the wheelbase can be calculated.

ISSN 1653-0187 / ISRN LTU-PB-EX--09/080--SE / NR 2009:080